ROAD: Regularized Optimal Affine Discriminant

ROAD is an R package that implements the Regularized Optimal Affine Discriminant (ROAD) algorithm for high-dimensional classification. This package is a translation of the MATLAB implementation described in the paper:

Fan, J., Feng, Y., & Tong, X. (2012). A ROAD to Classification in High Dimensional Space. PDF

The ROAD algorithm addresses the challenges of high-dimensional classification by incorporating covariance information and regularization, leading to improved classification accuracy.

---

📦 Installation

You can install the ROAD package directly from GitHub using the devtools package:

install.packages("devtools")  # Install devtools if you don't have it
devtools::install_github("ghorbani-alireza/ROAD")

🚀 Quick Start

Here’s a quick example to get you started with the ROAD package:

library(ROAD)

# Generate simulated data
sim_data <- simulate_road_data(p = 1000, n = 300, s0 = 10, rho = 0.5, randSeed = 1)

# Extract training and testing data
x <- sim_data$x
y <- sim_data$y
xtest <- sim_data$xtest
ytest <- sim_data$ytest

# Fit the ROAD model
fit <- road(x, y)

# Perform cross-validation
fit_cv <- roadCV(x, y, fit)

# Make predictions
predictions <- roadPredict(xtest, fit, fit_cv)

# Calculate test error
test_error <- mean(predictions$class != ytest)
cat("Test Error:", test_error, "\n")

---

🛠️ Features

High-Dimensional Classification: Handles datasets with a large number of features.

Regularization: Incorporates regularization to improve classification accuracy.

Cross-Validation: Includes tools for cross-validation to evaluate model performance.

Simulation Tools: Provides functions to generate simulated data for testing and demonstration.

📖 Documentation

For detailed documentation, check out the help pages for each function:

?simulate_road_data  # Generate simulated data
?road                # Fit the ROAD model
?roadCV              # Perform cross-validation
?roadPredict         # Make predictions

##📜 License

This package is licensed under the MIT License. See the LICENSE file for details.

##🙏 Acknowledgments

The original MATLAB implementation by Yang Feng. The authors of the paper for their groundbreaking work on high-dimensional classification.

📧 Contact

For questions, feedback, or contributions, please contact:

Alireza Ghorbani
Email: ghorbanialireza@outlook.com

Enjoy using ROAD! 🎉